March 4, 2002
3.1.1 - History and Physical Environment of the Petitcodiac - Denis Haché
- Sediment build-up began during causeway construction.
- Tidal bore was 2 m high in 1906, now 0.20 m high.
- Mud deposition and erosion fluctuate a great deal, even between tidal cycles.
- 9 m difference in mud elevations between freshets and low flow.
- Channel bottom about 10 m higher than historical value.
- Accumulation of mud, about 120 x 106 m3 to date.
- About 16 x 106 m3 have accumulated 1991-2001. Recent accumulations occur further downstream.
- About 10 x 106 m3 of sediment have accumulated in the head pond above the causeway.
- Many examples of gates being opened - 1968, '88, '89, '90, '98 and '99. - Can use information recorded during these events to help understand processes.
- Estuary is not in equilibrium - mud accumulation occurs further downstream and head pond continues to fill in.
Report reference: Feasibility of Modelling the Petitcodiac Estuary, Report No NBE-1 - David H. Willis and Associates Ltd, 31 March 1999
- Hydrodynamic models technically feasible, but not necessarily economically feasible - extensive field data required.
- Mud properties require more investigation. Modellers and sedimentologists have differences of opinion.
- Need a specific model for the Petitcodiac rather than using an existing commercial model.
- For the purpose of this exercise fluid mud is assumed where concentration are 10 mg/l.
3.2.1 - Modelling of Complex Flow Conditions - David Willis for Norm Crookshank
- Diverse modelling capabilities are available through the Canadian Hydraulic Centre at National Research Council of Canada.
- 1-, 2- and 3-dimensional spatial modelling capabilities.
- Each model has strengths and weaknesses. For example, Telemac model is good for incorporating wetting and drying of sediments. The CUMBSED model deals with erosion and deposition. Environmental Simulation System (EnSim) has a range of models and applications.
- Examples of applications - Athabasca River and Rideau River in Canada and Laem Chabang, Thailand.
3.2.2 - Modelling of Turbulence - Dr. Noboru Yonemitsu
Skeptical of numerical models because of:
- interaction of fluids and sediments,
- complexity of boundary conditions,
- mixing time ranges from 1 second to hundreds of days,
- estuaries are site-specific and strongly 3-dimensional,
- assumptions that are based on averages need to be challenged, and
- most estuaries are density stratified which inhibits vertical mixing.
3.2.3 - Finite Element Modelling - Dr. Dave Greenberg
- Finite element approaches are versatile and resolution can be increased in areas of interest.
- Experience in modelling to simulate blockage for tidal power in the upper Bay of Fundy shows that impacts of dams on tidal regime can be far reaching (as far as Boston in some cases).
- To model the Petitcodiac we need recent bathymetric data as the seabed is changing e.g. significant scouring recently found in upper Bay of Fundy.
- LIDAR is a valuable source of high-resolution elevation data.
3.2.4 - Effluent Dispersion - Jochen Schroer
- Simple models good for predicting local effluent mixing (Miramichi River).
- Can link simple near-field mixing models with a hydrodynamic model to look at dispersion.
- Dispersion further away difficult to predict with simple models.
- Moncton sewage has primary treatment, no diffuser, and a relatively high effluent volume.
- No water quality objective is in place for the Petitcodiac River/estuary.
3.2.5 - Overview of Model Approaches - Andrew Driscoll
- Every model has strengths and weaknesses.
- 1-D models are simplistic but versatile.
- Terms of Reference for the workshop are ambitious and, will require a suite of hybrid applications involving 1-, 2- and 3-D models.
3.2.6 - Panel Discussion - Hydrodynamics Session
- We can model hydrodynamics without sediments, but concerns exist about validity.
- Need to measure role of ice in sediment movement.
- Need to validate our models.
- Need for present day data such as LIDAR surveys.
- 1-D model may be appropriate for evaluating gate control.
- Waves need to be considered, especially in outer bay.
3.3.1 - Cohesive Sediment Transport - Dr. Krish Krishnappan
- A sediment model needs to predict concentration from surface to bottom.
- Only valid approach is to treat fluid and sediments in four separate layers: suspended, fluid mud, partially consolidated bed and fully consolidated bed.
- Fluxes between layers need to be specified.
- Shear stress controls size of flocs; when shear stress is low, fine particles stay suspended and heavier ones settle out.
- Instruments are available to measure shear stress and sediment stability.
3.3.2 - Sediment Behaviour in Turbid Environments - Tim Milligan
- Relationship between single grains and flocs depends on concentration and turbulence.
- Mud breeds mud.
- High concentration and varying shear stress,
- Rapid deposition, settling velocity of fluid muds approximately 1 cm/sec.
- Flocs and aggregates behave differently so require different models,
- We don't know the in-situ settling velocity of flocs,
- Stickiness of flocs is biologically determined and affects their size.
Implications for Petitcodiac:
3.3.3 - Properties of Near-Bottom Fine Sediment Suspensions - Dr. Richard Faas
- Settling time controls sediment density.
- Fluid mud does not behave like Newtonian fluids where shear stress and shear rate are constantly proportional.
- Fluid mud tends to be of three types that aren't Newtonian.
- Viscosity varies with time and density and you cannot use a single value.
3.3.4 - Sediment Budget Estimates - Dr. Carl Amos
- About 1 x 106 m3 of sediment in suspension at any time in the upper Bay of Fundy including Cumberland Basin and Shepody Bay.
- Causeway has acted like a vacuum cleaner and removed 120 x 106 m3 of sediment from Bay of Fundy that would have otherwise contributed to salt marshes downstream.
- Turbidity has to be measured in Petitcodiac system.
- Concentration of sediments is always higher in the inner bay.
- Vertical profiles have never been done in the Petitcodiac system.
- Red Petitcodiac sediments are excellent for satellite imagery.
- Tidal range is amplifying in Bay of Fundy - 3000 years ago Moncton was at the tidal limit.
- Effects of ice are unknown and are probably important.
- Sediment sources are 85% seabed, 12% cliff erosion, and 3% from rivers.
3.3.5 - Panel Discussion - Sediments Session
- Can measure average turbidity in Petitcodiac with satellite technology.
- For vertical distribution of concentration need to calibrate with in-situ samples.
- Muds accumulate in front of causeway because of lower shear stress (less turbulence).
- Sediments sucked out of the upper Bay of Fundy system and deposited near the causeway rather than creating wetlands elsewhere.
- Sediment concentration can be measured with Optical Backscatter (OBS).
- Vertical distributions of current can be measured with an Acoustic Doppler Current Profiler (ADCP) (in most environments).
- Should we undertake data mining to make full use of existing data?
- Need good measurements from the field e.g. density flocculation vs. time, channel depth, settling velocity.
- Winter observations are problematic and have traditionally been scarce.
- Make use of existing Gunningsville Bridge as a platform for measurements.
- Use LIDAR for channel bathymetry or drying area morphology.
- Remote sensing by satellites to measure average turbidities.
- Is the estuary in equilibrium?
- What is the mass balance of sediments in the system?
- What are the mud properties?
- Where is appropriate downstream boundary when modelling sediment fluxes?
- What is the fate of seasonally deposited sediment? Is it re-suspended?
Suggestions:
Fundamental questions remaining:
3.4.2 - 1B - What are the important parameters for modelling purposes? Facilitator - David Willis
- Issues for modelling in the Term of Reference for the Workshop are too detailed and should not be followed at first. Sufficient for two or three major steps in future.
- Revise Terms of Reference issues to take a more macro approach to develop basic understanding.
- Use box models and 1-D models to determine sediment budgets and movements (zone of erosion, deposition) under different scenarios.
- Take advantage of gates to do experiments, in conjunction with 1-D and box models to understand possible change.
- Need comprehensive measurements of the estuary: bathymetry, turbidity, morphology, sediment loads, biology of sediments (nitrogen, phosphorus, bacteria etc).
- Hydrodynamic master model should be built for all Maritimes to serve all necessary coastal applications throughout region. This would require a multi-partner approach.
- Detailed field data are needed to determine if there is vertical stratification due to salinity and sediment concentrations. (May have ramifications for 3-D modelling, which is also useful for shear stress calculations.)
- Field data are needed in other areas such as ice processes.
- Present modelling capabilities are inadequate for a single, all encompassing model (hydrodynamics coupled with sediment, ice processes etc.).
- More realistic to use a combination of models of different complexities.
- Certain aspects of the problem (e.g. motions, location of meanders) cannot be modelled. They have to be considered using empirical approaches.
- Models cannot predict future conditions, however they can be used to compare scenarios.
- Model development should go hand in hand with data acquisition; possibly a step-by-step process of modelling, calibration/testing, data acquisition, model improvements, more data etc.
- Is the potential restoration of the tidal bore important? Should the model aim to simulate it?
- Better identification of source(s) of sediment sequestered in estuary is needed.
- Extreme runoff events and associated sediment transport need to be considered; also the possibility of ice jamming in the estuary if the causeway is removed.
3.5.1 - Cohesive Sediment Transport - Dr. Erik Toorman
- There are many uncertainties when modelling. Physical and numerical models each have strengths and weaknesses. Data to support calibration and domain boundaries can also introduce uncertainty.
- Weaknesses in models are: how to model exchange of bottom sediments, high concentration effects, effects of laminar flow, bulk erosion and bedload transport.
- Application shown for Flyland Netherlands and Mont Saint-Michel, France. Latter used both physical and numerical models (used to define boundary conditions).
- We need to validate models under many conditions.
- Numerical models could not reproduce lab results even when complete data were available.
- Different models often do not compare well (engineering and research models).
3.5.2 - Estuarine Modelling Methods and Projects - Dr. Ole Petersen
- Estuary model have two parts, which interact with each other through diffusion, density and depth.
- Hydrodynamics = flow, turbulence, and salt/temperature.
- Sediments = suspended sediments and bed deposits.
- Modelling is a tool for rational discussions.
- Models provide a window to look at reality. Can move window around to change the view but expensive to increase its size to look at more factors.
- Different problems require different models. Need to choose the model most relevant for problem.
- Models require site-specific data.
- Examples of models: Tamar estuary (England), Mariager Fjord (Denmark), Loire River (France), Hamilton Air Base (San Francisco) and Belt Links (Denmark).
- Multi-disciplinary research is important.
- Petitcodiac River/estuary is perfect for collaborative studies such as the Littoral Investigation of Sediment Properties (LISP).
- All barriers are different; each has its own impacts on the sediment and hydrodynamic system.
- When we jump in and modify environments, we take a huge risk.
- On Evangeline Beach, Minas Basin, Nova Scotia the stability of sediments in summer is based on Corophium/diatom interaction.
- It took over 30 years to see effects of causeway constructed on Annapolis River, Nova Scotia.
- We should take the community on as partners.
- Scientists have a responsibility to explain what is being done and why.
March 5th, 2002
3.7.1 - Particle Trapping in Stratified Estuaries - Dr. Bob Chant
- Vertical stratification is an important parameter in an estuary.
- In flood-dominated estuaries (e.g. fast flood, slow ebb), particles are brought into the estuary.
- Tidal bore is an extreme example of a flood tide dominated system.
- Buoyancy effect on flood-dominated estuaries enhances the trapping of particles.
- Stratification turns off mixing on the ebb tide and particles remain trapped in the estuary.
3.7.2 - Storm Surges - Dr. Hal Ritchie
- We need a tidal gauge to develop a historical record to predict storm surges in the Petitcodiac River/estuary.
- Bay of Fundy storm surges are difficult to forecast because of non-linear interactions with high tides and lack of tidal gauge information.
- Research is required for improving accuracy of Bay of Fundy storm surge but it is feasible.
- The existing gauging station at Saint John Harbour is too far away and recently shown to have inaccuracies in its data record.
- Ice changes salt, discharge, current, temperature, and water quality of an estuary.
- Petitcodiac estuary has 5 zones: above the causeway normal river ice; Zone 2 vertical ice occurring causeway to Dover; Zone 3 and 4 where ice produced travels upstream; Zone 5 ice-free.
- Winter and summer conditions require different models for Petitcodiac.
- Modelling of ice conditions requires three components: ice generation, transport of drift ice and subsequent stranding.
- In mid-winter generated, transported and stranded volumes of ice depend on estuary width and length.
- During melting, need to model ice sediment interaction and evaluate sediment-laden ice.
- Changes in salt/freshwater volume changes heat input and hence the ice regime.
3.8.1 - Tides in Bay of Fundy - Charlie O'Reilly
- World's highest tides with a range of 16.3 m occur in Minas Basin, Nova Scotia.
- Bay of Fundy tidal ranges are increasing.
- Scouring of ocean bottom appears to be up to 40 m off Cape Enrage.
- Subdued perigean spring tides are predicted from 2003 to 2008 minimizing any impacts of storm surges.
- Most recent bathymetry survey for Petitcodiac estuary chart was in 1941.
- Datums are to be harmonized worldwide to lowest predicted tide NAD83.
- Datum "very unsure" at Moncton.
- Need to install tidal recording stations in the estuary to understand tidal characteristics.
3.8.2 - Hydrographic Survey of Petitcodiac - Jean-Claude Vautour
- Complete surveys of Petitcodiac are available from 1861, 1991, and 2001.
- Consist of 33 cross-sections of the river about 1.5 km apart.
- Done from causeway to Shepody Bay.
- Recent data are referenced to geodetic survey. The 1861 survey has an unknown reference.
- Vertical accuracy of recording depth sounders depends on sediment, speed of sound, tidal correction and transect position.
- Only 2001 data available in x, y, z digital format.
- No salinity profiles available.
3.8.3 - Remote Sensing - Tim Webster
- RADARSAT - gives land cover characteristics in all weather conditions, cost $650 per 100 km2.
- RADARSAT 2 - as above but uses vertical and horizontal beam.
- Ikonos - less than 1 m resolution, cost $39 per km2. Data cheaper from archive.
- QuickBird - 0.7 m resolution, cost $75 per km2.
- LIDAR - surface heights at 3 m point spacing and +/- 15 cm vertical accuracy, wet, smooth mud not a good target, need fair weather, cost minimum $20,000.
- CASI - programmable bands, need geometric corrections, cost $28 per km2.
3.8.4 - Data Acquisition in Estuaries - Gary Bugden
- Deployment and recovery of instruments is difficult in muddy environments.
- Acoustic Doppler Current Profiler (ADCP) can be oriented in up, down or horizontal positions. Problems with distortion of echo in freshwater layer and fluid mud layer.
- Flow though systems are available to measure sediment and salinity.
- Conductivity-temperature-depth (CTD) probes may become clogged with sediment.
- Need to test to see what instruments work in the Petitcodiac estuary.
- Don't have enough field data to even look at model design.
- Older data collection methods for more types of data may be best and can involve the general public.
3.8.5 - Panel Discussion - Data Collection Session
- High frequency depth sounder may show fluid mud in Petitcodiac. Low frequency may show the bottom. Difference between the two could give you fluid mud layer.
- High frequency Acoustic Doppler Current Profiler (ADCP) might work better than lower frequency instruments in the Petitcodiac.
- Bottom tracker combined with digital global positioning system could be used to measure mudflow.
- Gamma attenuance array can look at high turbidities, and has been used in Bay of Fundy and in Holland.
- Density stratification in Petitcodiac may be due to sediments and not salinity.
- Riverview Fire Department is willing to provide free use of boat for data collection.
- Datum uncertainty to which not only tidal water levels, but also bathymetry are measured is a source of modelling problems.
- The data you need depends on the model you are using.
- Need to collect and collate bathymetric data.
- Need water level measurements (tidal bore, hydrodynamics) from at least three points over time.
- Need to collate freshwater discharge information.
- Need short-term detailed measurements at key sites at key times.
- Also need long-term measurements at key sites to get seasonal variations of factors such as velocity and turbidity.
- Could put a camera on radio tower beacon to take time-lapse photographs of river. Little danger of losing or damaging equipment.
- Geometry of the channel will be affected by ice.
- Start simple!
- A barrier with a door is an opportunity not usually found in other estuaries.
- Some sort of experiment is required to: fill specific data/knowledge gaps (e.g. bank stability in the near-field), to do extensive data mining of collected information to further understanding, and to see if models are believable and support model decision-making.
- Need to know limitations that would be imposed on gates as part of the Environmental Impact Assessment, otherwise experiment likely to fail.
- Understanding of existing situation (sediments) should be done before any experiment. Need to understand the existing processes.
- If experiments are done, do extensive monitoring of all parameters because the information base is so limited.
- If Memramcook causeway gates were opened first, this could be used as a smaller scale test to monitor changes.
3.9.3 - 2C - How can data collection be optimized for modelling? Facilitator - Tim Milligan Process:
- Basic data.
- Conceptual model.
- Additional data.
- Refine model.
- Repeat 3 and 4.
- Data mining and fundamental data collection is required to develop conceptual models. Develop a step-wise model with feedback from ongoing data collection.
- For hydrodynamic models need: tide, current, channel profile (time dependent), boundary conditions, river discharge (flow conditions, water level), density structure (total suspended solids, salinity, temperature), synoptic measurements.
- Sediment dynamics: concentration, settling velocity, erosion.
- Temporal/spatial evolution of parameters: seasonal effects (ice, biology), sediment budget, close interaction with hydrodynamics.
- Use historic channel data to start a simple model?
3.10 - Case Studies Session - Other Experiences on Estuary Modelling
3.10.1 - Recent Experiences in Modelling Cohesive Sediment Transport - Dr. Doug Scott
- Can waste time on "battle of the modellers" arguing about who has best scenario. A model is only a tool to understand the processes. You also need to look at the field data in some detail.
- It is critical to have integrated team approach.
- Use more than one sediments testing facility when determining basic characteristics.
- Define boundary conditions (otherwise too much variability).
- Need to integrate models and measurements to come up with solutions.
- Changes to Petitcodiac estuary need to be controlled with ongoing monitoring, modelling, and evaluation.
3.10.2 - Erosion Modelling - Implications for the Petitcodiac - Dr. Michael Davis
- Models are needed - large and small-scale and 1- and 2-D for both hydrodynamics and morphological factors.
- Need to measure, experiment, and talk to people who know river and look at historical analysis (e.g. Bray papers).
- Do some quick and early field work (for example on erodability, characteristics, roughness, size of flocs etc.).
- Fieldwork is essential. Use Gunningsville Bridge, or work from shore near Chateau Moncton.
- Difficult to determine sediment characteristics in Petitcodiac estuary.
- Use data mining to set boundaries for flow and sediment.
- Publicize the data and the science to build credibility.
3A, 3B and 3C - What modelling approaches would be suitable for the Petitcodiac?
- Models can now be used to model hydrodynamics, wind, ice, gate operations, pollutants, but not sediment. An approach would be usually a 1-dimensional model leading to 2-dimensional horizontal model. Use finer grid in area of interest.
- To model sediment, erosion, deposition, transport, bank erosion, use a 2-dimensional vertical model leading to a 3-dimensional model.
- To model the tidal bore, a physical model (on site?) may be required. Numerical models of the bore will need to be at least 2-dimensional horizontal model.
- To model ice, a 2-dimensional horizontal model for ice transport and jams, and a conceptual model of winter processes including sediments are suggested.
- For prediction of long-term geomorphology, a 1-dimensional model may be considered. See paper on conceptual equilibrium in Bray, DeMerchant and Sullivan 1982.
- Numerical hydraulic models of the Petitcodiac will require a 'grid spacing' of 10 m to 25 m in the channels and mudflats of the area of interest, and 250 m to 300 m in Shepody Bay. Finite-difference models can achieve this by nesting
Need a model that is conceptually accurate.
Issues:
- What is role of suspended load in affecting fluid load and mixing?
- What is role of salinity stratification?
- What is role of tidal asymmetry?
- Can water property measurements be related to seasonal changes in channel?
- What is role of ice?
- What is role of biology?
Need to measure cross sections of channel, topography, total suspended sediments, density, salinity, velocity, erodability etc.
3.11.3 - 3C - How long should data collection and modelling take? Facilitator - Doug Bliss
Modelling approaches:
- Use smart people experienced with Petitcodiac problems.
- Start to measure things. Data will indicate conceptual models.
- Use gradual, step-wise approach to modelling.
3.12 - Summary of Workshop. Facilitators - Doug Bliss and Dr. Michael Chadwick
Something has to be done to increase the basic understanding of processes in the Petitcodiac River/estuary and we have to be smart about what we do. We need to stop speculating, because there is no basic information or understanding.
Common Themes of the Workshop:
- Poverty of basic information.
- Winter regime has different effects.
- Look at simple approaches first.
- Look at entire system from the headwaters to Bay of Fundy.
- An inter-disciplinary approach is key.
- Use the community for monitoring.
- The sky is the limit when it comes to modelling approaches (i.e. modelling approaches can go from the simple to very complex).
- Use model(s) appropriate for question(s) asked.
- Dedicate time and effort to mine existing data to increase understanding.
- Basic characteristics need to be extensively monitored during four seasons. Characteristics such as: temperature, salinity, turbidity, distribution of sediments, grain size, mud properties - shear, strength, consolidation of beds, nutrients, bathymetry, tides, water levels, ice distribution, biological inventories…
- Develop conceptual model first.
- Use simple model approach: 1-D, box, experiment, and physical.
- Evaluate and add levels of complexity if appropriate.
Note: The possible next steps stated below are:
- not binding on the Environmental Impact Assessment process,
- not of equal importance or value,
- have not been tested for validity, and
- have not been subject to formal evaluation.
- Create an inter-disciplinary science and engineering team.
- Develop a conceptual model of the estuary.
- Test to see if instruments (conductivity temperature depth probes (CTD), Optical Backscatter (OBS)) work in the muddy waters of Petitcodiac.
- Initiate field measurements of turbidity, temperature, salinity, and depth at two stations below causeway.
- Digitize aerial photos that define where the river has been.
- Create data bases of existing data and information.
- Create a web site on the Petitcodiac River/estuary and update information regularly to keep everyone informed.
- Figure out what needs to be done before resources are allocated for more data collection, interpretation, and modelling.
- Obtain tidal information at mouth of river. Costs to install another tidal gauge would be about $25,000, including other instruments to measure salinity and temperature.
- Look at good bathymetric survey data from Petitcodiac. Conduct another survey.
- Create an impartial evaluation committee to allocate money, give advice and to provide peer review. Could follow the example of Hudson River Foundation, www.hudsonriver.org.
- Invest some money with groups that could help with monitoring.
- Ensure workshop participants are contacted for advice on fieldwork and where to start.
In order to initiate work on the Petitcodiac, invest 10 % of budget for preliminary analysis and data collection. Items to be considered in the short-term could include the following:
On the longer term the following could be attempted: